Synopsis On the morning of 12 September 1995, the SIMCOE ISLANDER suddenly capsized while transporting a road vehicle loaded with large limestone rocks. The accident occurred in fair weather conditions when the cable ferry was approximately halfway across Boat Channel, proceeding from Wolfe Island to Simcoe Island. In the capsizing, the ferry operator fell overboard and the truck driver was carried to the channel bottom inside the cab of the discharged vehicle. Both persons surfaced and reboarded the overturned ferry, and were subsequently taken to Simcoe Island by a small pleasure craft. The Board determined that the cable ferry SIMCOE ISLANDER, which was heeled slightly to starboard and trimmed by the stern on departure, capsized because longitudinal and transverse stability were lost when the vehicle loaded on deck shifted. The initial backward movement of the truck was induced by the ferry motion and poor frictional contact between the vehicle's worn tires and the sloping wet surface of the wooden deck sheathing, and was not arrested because of ineffective rear wheel chocking and the absence of any other securing arrangement. 1.0 Factual Information 1.1 Particulars of the Vessel 1.1.1 Description of the Vessel The SIMCOE ISLANDER is a self-propelled cable ferry of all-welded steel construction. She was fitted with diesel-engine-driven hydraulically powered propulsion gear, friction-drive wheels, fairleads, and 19mm-diameter wire cable, which, together with the enclosed control shelter for the operator, are located on the starboard side of the main deck. The hull is of simple barge form with symmetrical raked ends, and is sub-divided into 12 watertight compartments by two longitudinal and three transverse bulkheads. Hydraulically controlled loading/unloading ramps are centrally located at each end of the ferry. The main deck is fitted with wooden sheathing, laid fore and aft on each side of the centre line in way of the vehicle wheel lanes, as shown on the Outline General Arrangement Plan (see AppendixA, Figure No.1, and AppendixC, Photograph No.1). The ferry guide cable is submerged and spans Boat Channel between Wolfe Island and Simcoe Island. The channel is approximately 480m wide and 7m deep at its midpoint. The ferry crossing normally takes between four and five minutes to complete. (See AppendixB for sketch of the area.) 1.2 History of the Voyage At about 1035(3), 12September 1995, the cable ferry SIMCOE ISLANDER, after loading a large truck carrying some eight long tons (hereinafter referred to as tons) of limestone boulders, began crossing from Wolfe Island to Simcoe Island. The weather conditions were relatively light, with a 15kph south-westerly wind, slight swell, and a 30cm chop acting on her port side. In preparation for departure, the truck driver drove onto the deck of the ferry, positioned the vehicle as directed by the ferry operator, turned off the ignition key leaving a low forward gear engaged, and applied the parking brake. The truck was parked symmetrically about the fore-and-aft centre line of the main deck, and the forward end of the truck's cargo box was abreast the forward end of the operator's shelter. After he had raised and secured the after loading ramp and ensured that the ferry was floating free of the shore, the operator placed wooden chocks behind the rear wheels of the vehicle. He then returned to the control shelter and applied forward hydraulic power to the guide cable. When under way, the ferry was heeled slightly to starboard due to the weight and hauling action of the guide cable, and also because of an accumulation of flood water in a leaking starboard side underdeck compartment. The ferry was equipped with a portable gasoline-driven bilge pump which had last been used to pump clear the same compartment shortly before loading the same vehicle about one and a half hours earlier that morning. The pump had not been employed during intervening crossings when the ferry was lightly loaded with either one or two automobiles, and was not in use during this second heavily loaded crossing. The ferry was also trimmed by the stern, partly due to the slight squat induced by the ferry's forward speed, but mainly because of the trimming effect of the heavy truck, the centre of gravity of which was aft of midships. The after freeboard of approximately 30cm observed on departure by the operator was similar to that on previous heavily loaded crossings, during which spray and water were routinely shipped on the after end of the main deck in similar weather conditions. The truck driver remained seated in the vehicle cab with both door windows open. As the ferry proceeded toward Simcoe Island, the operator left the control shelter to collect the fare, and remained standing on the port side of the deck, slightly forward of the truck. At about 1037, approximately halfway across the channel, both persons suddenly became aware that the truck had started moving toward the after end of the ferry. Neither the operator nor the truck driver felt any bump or other untoward movement of the vehicle, its load of rocks, or the ferry, nor did they notice whether any of the truck wheels were turning or sliding. The driver immediately applied the hydraulic foot brake; however, the backward movement of the truck continued to accelerate, and the momentum caused the ferry to trim markedly by the stern and to ship water on the after end of the main deck. As the truck continued toward the stern, the after half of the main deck became completely submerged and the ferry suddenly capsized, discharging the vehicle to starboard. The ferry operator fell outboard to port, while the truck driver was carried to the bottom of the channel inside the cab of the discharged vehicle. The operator quickly swam to and reboarded the overturned vessel. The operator subsequently located the truck driver, who had escaped from the submerged cab and had cleared the guide cable and starboard side rails on his way to the surface, and he assisted him on board. The sound of the capsizing drew the attention of nearby fishermen, who quickly went to the overturned ferry and transferred the survivors to Simcoe Island in their outboard-powered boat. The ferry operator and truck driver were startled by the rapidity with which the capsizing had occurred, and both subsequently estimated that the elapsed time between first becoming aware of the backward movement of the truck and being in the water was no more than five seconds. At about 1050, the ferry operator phoned the Wolfe Island municipal office and reported the occurrence. At 1055, the town clerk informed the Transport Canada (TC), Marine District Office at Kingston, which in turn informed Rescue Co-ordination Centre (RCC) Trenton, thereby initiating official Search and Rescue (SAR) response. TC Marine(4) also initiated pollution control response. The capsized ferry was hauled to Wolfe Island and subsequently righted at 1825, 13September. The vehicle was recovered and hauled to Simcoe Island on 15September. 1.3 Injuries to Persons The ferry operator and truck driver were not physically injured during the capsizing and did not seek medical attention, nor did they subsequently report any ill effects as a result of their immersion. Neither wore any personal flotation device, and the truck driver was a non-swimmer. 1.4 Damage 1.4.1 Damage to the Vessel Post-casualty examination of the righted ferry showed that the safety rails and control shelter were damaged during the righting operation. However, apart from extensive water immersion damage to the machinery and to the electrical and communications equipment, there was only superficial structural damage directly related to the capsizing. Pressure testing showed that the gaskets of several flush-fitting bolted access plates in the main deck were not watertight, and that some bilge/flood water found in otherwise intact underdeck compartments had entered while the deck was immersed and the ferry inverted. 1.4.2 Damage to the Environment The ferry diesel fuel and hydraulic oil tanks' total capacities are 191L and 168L respectively, and at the time of the occurrence, each tank was approximately 75percent full. The truck is reported to have had some 200L of gasoline fuel in its tank. Most of these oils escaped immediately after the capsizing and quickly dissipated in the atmosphere. The remaining oil slick was contained by pollution containment booms deployed by the Canadian Coast Guard (CCG) around the ferry and vehicle as they were being recovered and/or righted. The small quantity of spilled oil remaining was subsequently recovered from within the containment boom with oil absorbent material, for removal and controlled disposal. 1.5 Vessel and Crew Certification The vessel was in possession of a valid Ship Inspection Certificate (SIC)24 approved for Minor Waters ClassII Operation with 12 Passengers. Because of the location and restricted nature of the ferry operation, there is no mandatory requirement for operator certification; however, all of the operators hold valid TemporaryEngineer certificates, which are renewed annually by TC Marine. The operator in charge of the ferry at the time of the occurrence has held this qualification for 13 years. The TC publication entitled Stability, Subdivision, and Load Line Standards (TP7301) includes minimum stability standards that have to be met by vessels of different size and type for regulatory approval. However, since the ferry was constructed before 01June 1977, and as she is certificated to carry no more than 12 passengers, none of the standards apply to the SIMCOE ISLANDER. There is no record of the preparation or submission of any detailed as-built stability information. 1.6 Weather Information The weather at the time of the occurrence was clear, with a local south-westerly wind of approximately 15kph, causing a slight swell and a chop of about 30cm which, at the time of the occurrence, acted on the port side of the ferry. The reported wind and sea conditions were markedly less than those in which service is usually curtailed; however, during an earlier crossing on the morning of the occurrence, when the ferry was loaded and under way with the same vehicle on board carrying a similar load, the wind and wave action had been such that spray and water were shipped on the after end of the main deck. 1.7 Life-saving Equipment At the time of the occurrence, the life-saving appliances on board the ferry were in accordance with the approved requirements listed on the current CCG SIC24. The appliances comprised a 2-person emergency boat, one 12-person buoyant apparatus, 13 adult and 2children's approved lifejackets, and 2 lifebuoys with attached heaving lines. All of the life-saving appliances were recovered after the capsizing; however, because of the rapidity of the capsizing, none was used, and only the lifebuoys floated free. Post-casualty examination revealed that: the emergency boat and buoyant apparatus were not fitted with retro-reflective tapes; the buoyant apparatus was not painted with a high visibility colour; and the buoyant apparatus was lashed to the shipside with a securely knotted rope with no quick-release mechanism. 1.8 Ferry Operational History and Loading Practices The ferry has been in regular service since 1964 without any major incident. The ferry is crewed by one operator, and continuous daily service is maintained with one stand-by and three regular personnel. A work shift system ensures that individual working periods do not exceed 7.75 hours on weekdays and 10.5 hours on weekends, when traffic is less frequent. The ferry routinely carries full loads of either three automobiles and 12 passengers, or one heavy truck. Vehicles are loaded and positioned as directed by the operators. Reportedly, the operators always fit chocks of 10cm by 10cm hardwood, some 40cm long, behind the rear wheels of the heavy trucks. When heavy trucks are being loaded, the shoreward end of the ferry often grounds as the vehicle passes over the ferry ramp, causing significant damage and abrasion to the bottom shell plating, and also increased resistance when leaving the dock. In order to ensure better handling during the crossing and also reduce the grounding effect on arrival on the other side of the channel, heavier vehicles are routinely located so as to ensure that the ferry is trimmed slightly by the stern. However, the reduction in freeboard at the after end due to such trim makes the ferry more susceptible to shipping and retaining water on the after end of the main deck. Spray and water are regularly shipped and retained on deck in relatively moderate weather conditions, particularly when the ferry is carrying heavy road vehicles. Reportedly, this service is curtailed when the combined swell and waves due to south-westerly winds approach or exceed some 75cm in height. Because freeboard and trim are less affected when the ferry is loaded with lighter and more evenly distributed automobiles, that service is often maintained in slightly more severe weather conditions. No formal loading or stability information is provided for the guidance of the operators, nor is any maximum load limitation notice on display at the ferry landings. Other than the daily record of fares collection, no formal record or log of operating procedures is maintained, and the level of bilge water in any of the underdeck compartments is not routinely monitored. The transport of heavy loads of limestone rocks from Wolfe Island has recently increased due to the construction of a new breakwater on Simcoe Island. There is no public highway scale on Wolfe Island, nor is there a means to weigh the rock at the quarry, so the weight of rocks on each vehicle is not precisely known. Consequently, to determine the weight of the truck and its load, the ferry operators rely on visual comparison of the volume and configuration of each truckload with those previously carried, and on the driver's estimate of the load. The decision to sail is based upon this approximation and the prevailing weather conditions. The vehicle involved in the occurrence made several crossings the previous day with another driver and reportedly similar loads, until the ferry operator curtailed service in the late afternoon due to a pronounced south-westerly swell exceeding 75cm in height. The truck driver at the time of the capsizing was making his first unaccompanied crossing after vehicle familiarization and one introductory trip with the other driver earlier that morning. 1.9 Vehicle Condition at the Time of the Occurrence The vehicle is a 1978 International Loadstar 1800 three-axle truck with a tare weight of 7,020kg and a rated gross weight of 22,800kg. At the time of the occurrence, the truck was reportedly loaded with 8tons (8,130kg) of limestone boulders, for a total laden weight of some 15tons (15,150kg). The cargo box of the truck was approximately 4.1m long, 2.4mwide, and 0.9m high, and the tailgate was removed to facilitate dumping the boulders, which were loaded toward its forward end and protruded about 0.5m above its sides. A diver's examination of the submerged vehicle during the recovery operation confirmed that the ignition key was in place and the ignition turned off, a low forward gear was engaged, and the parking brake, which mechanically locks the main drive shaft, was fully applied. After the vehicle was recovered from the channel bottom on 15September, examination showed only slight structural damage. The hydraulic pressure hose to the front right wheel brake was found to be severed. However, this damage reportedly occurred during the recovery operation, at which time the hose was temporarily sealed with a vice-grip wrench to prevent further loss of hydraulic fluid from the braking system. A provincial Ministry of Transportation inspection on 27September showed that 7 of the vehicle's 10 tires were in a worn but regulatory satisfactory condition, and that the treads of the two right rear outside tires were worn down below regulatory minimum standards. Furthermore, the tread of one of the left rear outside tires was damaged, exposing internal reinforcement wires over an area about 20cm long by 6cm wide. The abraded and rusty surfaces of the exposed wires were consistent with this damage having been incurred some time before the immersion of the vehicle in this occurrence. (See AppendixC, PhotographsNo.2 to No.4.) 1.10 Ferry Condition at the Time of the Occurrence There have been several modifications and additions to the ferry during her 31 years of service, and their accumulated extra weight has reduced the mean freeboard. The principal modifications include the replacement of the drive system with diesel-engine-driven hydraulic hauling gear and fairleads; the installation of larger loading ramps, lifting posts and counter weights; the fitting of additional compensatory trimming ballast; the provision of additional life-saving appliances and anchoring equipment, etc.; and the welding of chafing and reinforcement steel doubling plates to the bottom shell plating. Post-casualty inspection of the ferry showed that the No.1 starboard side underdeck compartment was flooded because of grounding and abrasion damage to the bottom shell plating. This previously incurred damage was unrepaired at the time of the capsizing, and a portable gasoline-engine-driven pump had reportedly been in use for several weeks to intermittently pump the damaged compartment free of flood water. Examination of the wooden deck sheathing surface showed that diesel and hydraulic oil that had inadvertently spilled during routine refuelling and servicing, together with that dripped by previously loaded vehicles, etc., had been absorbed in areas adjacent to the propulsion engine, hydraulic drive, and loading ramp operating systems on the starboard side of the main deck. Consequently, the frictional characteristics of the sheathing surface were not uniform, and some areas were more adversely affected than others--particularly when wet with rain, spray, or shipped water. 1.11 Ferry Trim and Stability In view of the reported capsizing sequence, post-casualty stability examination focused on establishing the departure loading condition and the detrimental effects on trim and stability of the shipping of water on the main deck and the sudden backward movement of the vehicle. The post-casualty calculations are based on static conditions; however, the momentum of the moving vehicle would have caused dynamic changes of trim and main deck immersion significantly greater than those indicated. TC Marine records of construction plans and hydrostatic data approved when the ferry was built in 1964 indicate a lightship mean draught of 1'5 (0.43m), and a deck load of 12.95tons at a maximum mean draught of 2'1 (0.635m). The data do not indicate whether the lightship draught shown was a preliminary design draught or if it was verified after construction was completed. However, as the vessel was re-powered and several other modifications and additions are known to have been carried out during the past 31 years of service, the original lightship weight has increased. Consequently, the post-casualty calculations of trim and stability characteristics were based on hydrostatics and lightship details generated from hull measurements and a draught survey carried out before the ferry returned to service on 27September 1995. They indicated a current lightship mean draught of 1' 7 1/2 (0.495m). A post-casualty analysis/examination(5) of the vessel's trim and stability at the time of the accident showed the following: On departure, the ferry had relatively high initial stability, with a transverse metacentric height (GMt) of 2.46m, and a mean draught of 0.736m. The restraint of the hauling cable and the flood water known to have accumulated in the damaged No.1 starboard side compartment caused the ferry to be heeled some two degrees to starboard, and the fore-and-aft location of the vehicle caused the ferry to trim 0.184m by the stern. The combined effects of the aft trim and starboard heel were such that the static freeboard at the after end of the main deck centre line was 0.31m, and 0.222m at the after starboard corner. The after freeboard was reduced further by the slight settling of the stern when the ferry was under way, and the prevailing 30cm-high waves caused water to be regularly shipped on the after end of the main deck, particularly on the lower starboard side. When the moving vehicle reached the after end of the main deck, the static trim increased to 0.711m, and the after end centre-line freeboard was reduced to zero. The initial heel to starboard resulted in the after end starboard corner being immersed 0.095m. However, due to the wave action and the dynamic trimming effects of the moving vehicle, the effective immersion of the after end of the main deck was actually greater. With the vehicle 30cm off-centre, the static angle of heel would be such that the starboard after corner of the main deck would be submerged some 20cm.